Prolonged Drug Release from Gels, using Catanionic Mixtures

Full text

(1)Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Pharmacy 65. Prolonged Drug Release from Gels, using Catanionic Mixtures TOBIAS BRAMER. ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2007. ISSN 1651-6192 ISBN 978-91-554-7017-3 urn:nbn:se:uu:diva-8303.

(2)

(3)

(4)

(5)

(6)

(7) .

(8) ! " ##" #$%&' ( ) ( ( *) ) +! ( *) ,- .)

(9)

(10) /

(11)

(12) 0/) .- ##"- *

(13) 1 ( 2

(14) .

(15)

(16) - 3.

(17)

(18) -

(19)

(20)

(21)

(22)

(23) 4'- 44 - - 506 $"78$&8''8"#&"8 .) ( .

(24)

(25) 8(

(26) /

(27)

(28) ((

(29)

(30) ) (

(31)

(32)

(33) ( - 5 / ) /

(34) )

(35)

(36) ( .

(37)

(38) ) / ) ) (

(39) .) )

(40)

(41) ) ) /)

(42) ) .

(43) / 89

(44) ) / (

(45) - .) : ( ) 9 )

(46)

(47) (

(48)

(49) ) (

(50)

(51)

(52) ; / ) / (( ) ) ) - <)

(53) ) ((

(54) ((

(55) ( ( ) (

(56)

(57) .

(58)

(59) / )

(60) ( )

(61)

(62)

(63) ) ((

(64) / &# &## .) (( ( )

(65)

(66) )

(67)

(68)

(69) )

(70) ) .

(71)

(72) /

(73)

(74) ) ) (

(75)

(76) ) / (

(77) ;

(78)

(79)

(80) )- 3 ) ) ) ) ) / /) ((

(81) / (

(82) - .) /

(83) (

(84)

(85)

(86)

(87) )

(88)

(89) ) ( "-

(90)

(91)

(92)

(93) #-$= 6 0(

(94) ( /

(95)

(96) (

(97) ( )

(98) (

(99) ) ( +00,

(100)

(101)

(102)

(103) / ) 0 ) )

(104) ( ) )

(105)

(106)

(107)

(108)

(109)

(110)

(111)

(112) ( )

(113) )

(114) ) - 5

(115) ) )

(116)

(117)

(118)

(119) ) )

(120)

(121) ( ) /

(122) (

(123) - >

(124)

(125) ( )

(126)

(127) (

(128) ( .

(129)

(130) / ) ) ) (

(131) .

(132)

(133)

(134) ((

(135) (

(136) ! "

(137) #

(138)

(139) # " $ %&'# # ()*%+,- # ? . ##" 5006 &4'&84&$ 506 $"78$&8''8"#&"8

(140) %

(141)

(142) %%% 87 # +) %@@

(143) -9-@ A

(144) B

(145) %

(146)

(147) %%% 87 # ,.

(148) EVEN IF THE VOICES AREN’T REAL, THEY HAVE PRETTY GOOD IDEAS.

(149)

(150) CONTENTS 1. INTRODUCTION ...............................................................................11 1.1 GELS AS PHARMACEUTICAL DOSAGE FORMS ............................11 1.1.1 What is a gel?.........................................................................11 1.1.2 A dosage form of great potential?..........................................12 1.1.3 The advantages of using a gel................................................13 1.2 THE PROBLEM.............................................................................14. 2. AIMS OF THE THESIS......................................................................15. 3. CATANIONIC MIXTURES ...............................................................16 3.1 WHAT IS A CATANIONIC MIXTURE?...........................................16 3.2 CATANIONIC MICELLES ..............................................................18 3.3 CATANIONIC VESICLES...............................................................18 3.4 CATANIONIC MIXTURES IN A PHYSIOLOGICAL ENVIRONMENT............................................................................19. 4. EXPERIMENTAL SECTION.............................................................21 4.1 MATERIALS.................................................................................21 4.2 COMPOSING THE PHASE DIAGRAMS ...........................................23 4.3 PREPARATION OF GELS...............................................................23 4.4 RHEOLOGICAL MEASUREMENTS ................................................24 4.5 DRUG RELEASE STUDIES ............................................................25 4.6 DETERMINATION OF THE CMC...................................................27 4.7 THE INTERACTION PARAMETER .................................................27 4.8 STATISTICAL ANALYSIS .............................................................28. 5. RESULTS AND DISCUSSION..........................................................29 5.1 DO DRUGS FORM CATANIONIC AGGREGATES WITH OPPOSITELY CHARGED SURFACTANTS?.....................................29 5.2 DO CATANIONIC AGGREGATES AFFECT THE DRUG RELEASE FROM GELS? ...............................................................................33 5.3 WILL THE USE OF CATANIONIC MIXTURES WORK IN A PHYSIOLOGICAL ENVIRONMENT? ..............................................37 5.4 CAN THE OPPOSITELY CHARGED SURFACTANT BE CHANGED INTO SOMETHING LESS TOXIC? ..................................................42 5.5 SOME DISCOVERIES CONCERNING THE RELEASE MECHANISMS..............................................................................46 5.5.1 Electrodynamic investigations...............................................46 5.5.2 Regular solution theory..........................................................49.

(151) 6. CONCLUDING REMARKS ......................................................................53. 7. KONTROLLERAD FRISÄTTNING FRÅN GELER – SAMMANFATTNING PÅ SVENSKA .............................................55. 8. ACKNOWLEDGEMENTS ........................................................................57. 9. REFERENCES ........................................................................................59.

(152) PAPERS DISCUSSED. This thesis is based on the following papers, which are referred to in the text by the Roman numerals assigned below: I. Bramer, T., M. Paulsson, K. Edwards and K. Edsman. Catanionic drug-surfactant mixtures: Phase behavior and sustained release from gels." Pharmaceutical Research (2003), 20(10): 1661-1667. Reproduced with permission. ©2003 Springer. II. Bramer, T., N. Dew and K. Edsman. Catanionic mixtures involving a drug: A rather general concept that can be utilized for prolonged drug release from gels. Journal of Pharmaceutical Sciences (2006), 95(4): 769-780. Reprinted with permission of Wiley-Liss, Inc. a subsidiary of Johan Wiley & Sons, Inc. ©2006 Wiley-Liss, Inc and the American Pharmacists Association.. III. Bramer, T., G. Karlsson, K. Edwards and K. Edsman. Effects of pH and ionic strength on catanionic drug-surfactant mixtures used for prolonged release from gels. Journal of Drug Delivery Science and Technology (2007), 17(4): 285-291. Reproduced with permission. ©2007 Editions de Santé.. IV. Dew, N., T. Bramer and K. Edsman. Catanionic mixtures formed from drugs and lauric or capric acid enable prolonged release from gels. Submitted. V. Brohede, U., T. Bramer, K. Edsman and M. Strömme. Electrodynamic investigations of ion transport and structural properties in drug-containing gels: Dielectric spectroscopy and transient current measurements on catanionic carbopol systems. Journal of Physical Chemistry B (2005), 109(32): 15250-15255. Reproduced with permission. ©2005 American Chemical Society. VI. Bramer, T., G. Frenning, J. Gråsjö, K. Edsman, M. Bergström and P. Hansson. Implications of regular solution theory on the release mechanism of catanionic mixtures from gels. Manuscript..

(153) My contribution to the above papers was as follows: I.-IV.. I was involved in all parts, except for the cryo-TEM measurements.. V.. I was involved in the question formulation, interpretation of the data and the writing process, but not in performing the measurements or the calculations.. VI.. I was involved in all parts, except for development of the mathematical models used..

(154) ABBREVIATIONS. Alp E BAC C940 C1342 CAC CI CMC CTAB C6TAB C8TAB C12TAB CTAT D Da DDAB Dif DoTAC G’ G’’ Ibu Lid LPC NaOA Orph PAA Prop s.d. SD SDS SDBS SOS TTAB Tet. Alprenolol The interaction parameter Benzalkonium chloride Carbopol 940, cross-linked PAA Carbopol 1342, cross-linked PAA with lipophilic modification Critical aggregation concentration Confidence interval Critical micelle concentration Cetyltrimethylammonium bromide Hexyltrimethylammonium chloride Octyltrimethylammonium bromide Dodecyltrimethylammonium bromide Cetyltrimethylammonium tosylate The phase angle Diffusion coefficient Dalton (mass unit) Didodecyl dimethylammonium bromide Diphenhydramine Dodecyltrimethylammonium chloride The elastic (storage) modulus The viscous (loss) modulus Ibuprofen Lidocaine Lauryl pyridinium chloride Sodium oleate Orphenadrine Poly(acrylic acid) Propranolol Standard deviation Sodium dodecanoate Sodium dodecyl sulfate Sodium dodecylbenzene sulfonate Sodium octyl sulfate Tetradecyltrimethylammonium bromide Tetracaine.

(155)

(156) Introduction. 1 INTRODUCTION. 1.1 GELS AS PHARMACEUTICAL DOSAGE FORMS 1.1.1 What is a gel? Thomas Graham can be remembered for several of his achievements. For example, he was the very last to uphold the position of “Master of the Mint” in England, an office once held by Sir Isaac Newton, but which, after the death of Graham (in 1869) became a subsidiary office of the Chancellor of the Exchequer. Amongst chemists he is known as the father of colloid chemistry [1] and, along with other significant contributions, he conducted groundbreaking research in the area of diffusion [2]. Moreover, apart from being responsible for the coinage at the Royal Mint, in his research on the properties of silicic acid, Thomas Graham also coined the term “gel” [3]. Ever since the days of Thomas Graham there have been numerous efforts to define what a gel is, and yet it is doubtful whether any of the proposed definitions actually covers all materials falling in this category [4-6]. In 1926 Dorothy Jordan Lloyd wrote “The colloidal condition, the gel, is one which is easier to recognize than to define” [7], a statement that, with every new definition put forward, seems more and more apt. In 1949 Bugenberg de Jong defined a gel as a colloidal system of solid character, in which the colloidal particles somehow constitute a coherent structure, the latter being interpenetrated by a (usually) liquid system [8]. This was evolved by Hermans into the more exact “a coherent colloid disperse system of at least two components that exhibit mechanical properties characteristic of the solid state and that (all) the components extend themselves continuously throughout the whole system” [5]. To further add to the confusion, when examining the literature, the reader also needs to consider the term “hydrogel”. By definition, a hydrogel is composed of three-dimensional, hydrophilic, polymeric networks and large amounts of water (from Gr. hydro = water) [9-12]. A gel is not necessarily the same as a hydrogel, for example there are also air-containing “aerogels” 11.

(157) Introduction and oil-containing “organogels”, but since these examples are exceptions, the literature generally uses the terms “gel” and “hydrogel” synonymously [12]. As yet, no definition has been determined that satisfies all needs. In this thesis a practical definition is generally preferred: A rheological definition of the gel is used, stating that the elastic modulus (G’), also known as the storage modulus, dominates over the viscous modulus (G’’), also known as the loss modulus, and that G’ is frequency independent [13, 14], Figure 1. This definition is not all technical, however, Almdal et al also points out that the term gel should be limited to systems that consist of two or more components, one of which is a liquid, and that are soft, solid or solid-like [14]. 1,00E+03. G', G''. G’. 1,00E+02. 1,00E+01. 1,00E+00 0,001. G’’. 0,01. 0,1. 1. 10. Fq Figure 1. A typical example of a sample fulfilling the rheological definition of a gel; G’ is much higher than G’’ and G’ is stable over the entire frequency range. This particular measurement was performed on a 1% C940 gel.. 1.1.2 A dosage form of great potential? The all dominating form of pharmaceutical administration is, of course, the tablet, with about 4500 approved forms of administration currently having been registered in Sweden [15]. Other common administration forms are, for example, the capsule and liquid for injection, with about 800 and 1400 approvals having been granted respectively. The gel has a corresponding figure of only about 60 approvals. However, in Sweden, this figure still places it in the top third of the different forms of administration being used. Gels have been used for both local and systemic drug delivery, see for example the review by Peppas et al [11], and for numerous different routes of administration, for example for rectal administration [16-19], for vaginal 12.

(158) Introduction administration [20-22], as vehicles for iontophoresis [23-25], for nasal administration [26-28], for cutaneous [29-32] and subcutaneous [33-36] administration, for delivery to the stomach [37-40], for delivery to the colon [41-43] or for buccal administration [44-46]. In addition, numerous variations of gel-based controlled release administrations have been considered, as reviewed by Lin and Metters [47].. 1.1.3 The advantages of using a gel When a drug formulation is administered either on the skin or the mucosa, ensuring a long residence time at the site of adsorption is of importance. For example, a tablet would obviously not be suitable for application to the skin. On the other hand, in many cases a solution would soon dry out or would be mechanically removed before much of the drug had been able to pass any physical barrier. A typical example of this is a solution for ocular drug delivery, where the bioavailability is typically no more than about 1% [48]. The poor bioavailability is caused by the poor drug permeability of the cornea, in combination with nasolacrimal drainage. Due to some particular beneficial properties, to be described shortly, the gel has been suggested as a way of increasing the ocular bioavailability [49-54]. The rheological properties of a gel can be adjusted to prevent it from being mechanically removed, which probably is of particular importance for, e.g., ophthalmic formulations [11, 55, 56], but also when the intention is to target other areas such as the colon [41]. In addition, gels have muco- and bioadhesive properties, offering the possibility of creating an intimate and prolonged contact time at the site of administration [57-60], thereby improving the efficiency of a drug. The exact mechanism of bio- and mucoadhesion is unknown and can be presumed to vary with the choice of gel. However, adhesion is usually described as a two-step process, see for example the review by Edsman and Hägerström [57], where the first step is the creation of intimate contact between the dosage form and the mucosa, which arises as a result of the wetting and spreading of the gel. In the second step, interpenetration of the components occurs, through which the interpenetrating chains of the polymer and the mucosa can form weak chemical bonds with their environment. The combination of the rheological and muco-/bioadhesive properties allows the formulation of a drug vehicle that remains at the site of adsorption for an extensive period of time, improving the efficiency of the drug. Another advantage is the possibility of reducing the number of doses required over any given period, increasing the patients’ compliance.. 13.

(159) Introduction. 1.2 THE PROBLEM Even though there are great advantages to using gel as a pharmaceutical dosage form, there is one major problem that usually needs to be overcome. As a gel usually consists of more than 95% and often about 99% water, any single molecule within the gel will experience an environment that is very similar to pure water. This means that the drug molecule can diffuse readily and almost without interference through the gel, so the drug release from the gel is often rapid. Now, consider this feature together with the beneficial properties of the gel, i.e. its rheological and mucoadhesive properties, and it is evident that the rapid drug release feature makes the rheological and mucoadhesive ability of the gel quite useless. Hence, adopting a strategy to prolong the drug release is often necessary to obtain a fully efficient pharmaceutical gel. Several ways of obtaining a prolonged release from gels have been suggested, for example: x Suspending the drug as particles inside the gel [61]. x Letting the drug interact with the gel polymer [62]. x Distributing the drug to liposomes [63]. x Distributing the drug to micelles [64]. Obviously, there are problems linked to all of these methods. Suspensions of a drug requires that the drug is not too soluble, but yet is sufficiently soluble to be released within the required period of time; interactions with the polymer must be reversible and not affect the rheology of the gel too much and a poor integrity of any liposomes will affect the rate of drug release. This thesis therefore suggests a novel method of prolonging the drug release from gels, namely by the use of catanionic mixtures.. 14.

(160) Aims of the Thesis. 2 AIMS OF THE THESIS. The overall aim of this thesis was to explore catanionic drug-surfactant mixtures to investigate their potential as a novel way of obtaining prolonged drug release from gels. The specific goals were: ¾ To investigate how common it is for a drug to be able to be incorporated in catanionic mixtures. ¾ To investigate whether catanionic mixtures may be used as a way of prolonging drug release from gels. ¾ To study the effects of a physiological environment on drug-surfactant mixtures. ¾ To study the possibility of varying the oppositely charged surfactant mixed with the drug, in order to lessen the toxicity of the mixture. ¾ To study the release mechanism from the catanionic mixtures, from gels.. 15.

(161) Catanionic Mixtures. 3 CATANIONIC MIXTURES. The term catanionic mixtures was first used in 1987 by Jokela et al [65] in a study reporting swollen lamellar phases in four mixtures of various oppositely charged surfactants. A couple of years later, Kaler et al reported on spontaneous vesicle formation in aqueous mixtures of positively charged cetyltrimethylammonium tosylate (CTAT) and negatively charged sodium dodecylbenzene sulfonate (SDBS) [66]. Mixtures of surfactants had already been studied extensively over the years, including a small number of early studies on mixtures of oppositely charged surfactants [67-71], but it is not until the last 20 years or so that catanionic mixtures have really attracted attention and interest in these mixtures is still growing.. 3.1 WHAT IS A CATANIONIC MIXTURE? The word “catanionic mixture” is self-explanatory; it means a mixture of a cation and an anion. However, what the term fails to indicate is that both of the components additionally possess surface-active properties. Hence, there are two forces driving the formation of catanionic aggregates, arising from electrostatic and hydrophobic interactions. +. -. + A. B. Figure 2. Schematic of the catanionic mixture, where formation of (A) vesicles and (B) spherical, elongated and branched micelles can occur.. 16.

(162) Catanionic Mixtures. In Figure 3 a classic example of the diluted region of a catanionic mixture is shown, in which a micelle phase and a vesicle phase surrounded by a multiphase region, are displayed on each side of the phase diagram, separated by precipitation. This symmetric appearance is found in several studies published on catanionic mixtures, for example in the already mentioned study by Kaler et al on mixtures of CTAT/SDBS [66], as well as in mixtures of DoTAC/SD [72] and SDS/DDAB [73, 74]. However, in a review of this field, Gradzielski notes how the symmetry of the phase diagram is dependant on the differences in chain lengths of the two surfactants [75] — the larger the difference in chain length, the greater the asymmetry in the phase diagram. In addition, the type of structures formed within each system is dependant on both the cationic/anionic surfactant ratio and on the total concentration of both surfactants together.. Figure 3. A model of a classical catanionic phase diagram, reproduced from Bramer et al [76], where L represents a micellar/aqueous solution, V+ represents positively charged vesicles, V- represents negatively charged vesicles and P represents precipitate.. 17.

(163) Catanionic Mixtures. 3.2 CATANIONIC MICELLES Ionic surfactants commonly form spherical micelles when they are dissolved in water. The formation of micelles leads to a gain in free energy as the contact area between the water and hydrocarbons is reduced [77]. In the catanionic mixtures, however, not only is it common that various other structures form, but the micelles themselves often adopt new forms. The micelles within a micellar phase in a catanionic phase-diagram commonly shift between spherical (globular), to elongated (worm-like) and branched micelles (see Figure 2) as the total concentration [78] and the cationic/anionic ratio [79, 80] is varied. The influence of the surfactant tail length on the micellar size has been studied in mixtures of NaOA/CnTAB [81]. The largest micellar growth apparently occurs when one of the tails is long and the other moderately long (NaOA/C8TAB). This can be explained by the interactive forces being strong enough to induce a dramatic micellar growth, though not strong enough to induce bilayer formation or phase-separation. Mixtures comprising equal tail lengths (NaOA/C12TAB) or a long tail plus a very short one (NaOA/C6TAB), however, lead to phase separation or only weak micellar growth respectively. An interesting aspect of this study is how rheology was successfully used as a measure of micellar growth, which is something that has also been shown by, for example, Koehler et al [82]. In addition, it has been demonstrated that, when small amounts of anionic SDBS were added to cationic CTAT micellar growth was initially promoted, followed by a structural change: The viscosity increased gradually until it reached a maximum value whereupon it decreased again. This was interpreted either as the elongated micelles having become less rigid, or there could have been a transition to branched micelles of reduced viscosity [82].. 3.3 CATANIONIC VESICLES Traditionally, several different methods are used for the preparation of vesicles, such as sonication, thin-film hydration and high-pressure extrusion, whereas it under certain circumstances may be sufficient to vortex or just shake a mixture vigorously to accomplish vesicle formation [75]. Thus, one of the particularly interesting features of catanionic vesicles is their ability to form spontaneously. Kaler et al were the first to ascribe catanionic vesicles this property in a study on three different catanionic systems published in 1989 [66], since when this has been supported in numerous other studies and for numerous other catanionic systems [72-75, 83-90].. 18.

(164) Catanionic Mixtures The concept of the vesicle as a thermodynamically equilibrated state has been discussed more thoroughly than spontaneous formation. For example, Laughlin [91] dedicated an article entirely to this subject where he states that from a thermodynamical point of view, vesicles are not to be considered to be an equilibrium state. To illustrate the ongoing debate, Almgren and Rangelov [92] and Marques [93] studied catanionic mixtures of CTAB/SOS and DDAB/SDS respectively, to investigate the effects of the preaparation procedure. Both studies lead to the conclusion that the size of the catanionic vesicles formed was dependant on the preparation procedure. However, whereas, according to Marques this is proof of the vesicle being a thermodynamical equilibrium state, Almgren and Rangelov argued that it is proof of the exact opposite. However interesting this discussion may be, it is fortunately of limited value for most potential applications. Clearly, vesicles are readily formed in numerous catanionic systems and they appear to possess long-term stability qualities. There is a vast variation in the size and polydispersity of the catanionic vesicles within each system, which depends on the cationic/anionic molar ratio, as well as between different systems. Typical catanionic vesicles, as far as one can talk of such things, span a size of a couple of hundred nm.However, vesicles as small as 20 nm [73, 74, 87, 94-96] are commonly found, as are vesicles with a diameter of at least several μm [73, 96-98].. 3.4 CATANIONIC MIXTURES IN A PHYSIOLOGICAL ENVIRONMENT Several different approaches to the pharmaceutical application of catanionic mixtures have been suggested, as recently reviewed [76]. However, when considering the potential pharmaceutical applications, the physiological prerequisites must be fully appreciated. The physiological environment can usually be described by an osmolality corresponding to 0.9% NaCl and a neutral pH, but obviously the exact environment is utterly dependent on where the formulation is applied. It is conceivable that both the pH and the ionic strength in its immediate environment might affect a catanionic mixture. For example, it is common knowledge that salts alter the critical micelle concentration of various surfactants [99]. It has also been demonstrated that the size and shape of the micelles can be affected by the salt screening the repulsion between charged headgroups [81, 100-103]. Similar effects have been observed for catanionic mixtures. For example, in a catanionic system of CTAB and SOS, Brasher et al have shown that the addition of sodium bromide to the mixture can cause a vesicle-to-micelle transition [104]. 19.

(165) Catanionic Mixtures. On the other hand, it seems that catanionic mixtures are quite resistant to changes in temperature. For example, in a study on an octylamine/octanoic acid catanionic system, it has been shown that the phase extensions are only slightly affected by changes in temperature at physiologically relevant levels [105]. Similar results were obtained in a study on catanionic mixtures of alkyltriethylammonium bromide and sodium alkylsulfonate, where, in addition, it was concluded that neither the polydispersity nor the radius of the aggregates formed was much affected by changes in the temperature [106].. 20.

(166) Experimental Section. 4 EXPERIMENTAL SECTION. 4.1 MATERIALS The oppositely charged surfactants used in the catanionic mixtures examined in this thesis are shown in Figure 4 and the drugs employed are shown in Figure 5. Information on suppliers is provided within each manuscript and is not repeated here. The gels were composed of either Agar-agar or Carbopol 940. Carbopol is a poly(acrylic acid) covalently linked synthetic hydrogel [107]. When the pH is above 6 the polymer is highly charged and swells in water, forming a high-viscosity gel Benzalkonium Chloride (BAC) Cl Cl Lauryl pyridinium chloride (LPC) H3C R + + N N. CH3. Na Sodium lauryl sulfate (SDS). O. +. O S. O. O. Trimethyl ammonium bromide (TTAB) N. O. Na. Capric Acid. +. Br. +. O. Lauric Acid. O. Na. +. O. Figure 4. Structures of the oppositely charged surfactants used in this thesis.. 21.

(167) Experimental Section. The agar gel is uncharged and composed of polysaccharide complexes, originating from the agarocytes of several species of algae [108]. It is insoluble in water at room temperate, but dissolves upon heating. Upon cooling, a stiff gelatinous gel is formed.. Alprenolol. Diphenhydramine a: I-VI. O. N H a: II, IV. OH. b: 9.0. N. O. c: 3.7. b: 9.2 c: 2.9. CH3 CH3. CH3. Ibuprofen. CH3. a: II. CH3 O. a: II, IV b: 8.5. CH3. b: 4.4. H3C. H N. COOH. Lidocaine. N. c: 2.4. c: 3.7. Naproxen H. a: II. CH3O. CH3. COOH. Orphenadrine a: II, IV. O. CH3. b: 8.7 c: 4.1. N. b: 4.8 c: 3.0. OH O. Amitriptyline. H N. CH3. a: I b: 9.4. Propranolol. CH3. c: 4.9. a: II, IV. N. b: 9.1 c: 3.1. H2N. O O H3C. N H. Tetracaine a: I, III, IV, VI b: 8.5 c: 3.6. O. Atenolol. N(CH3)2 N H. O OH. a: IV b: 9.2 c: 0.1. Figure 5. Structures of the drug compounds studied, accompanied by information on: (a) which paper(s) the drug is used in (given in Roman numerals), (b) the pKa and (c) the log P of the drugs [109].. 22.

(168) Experimental Section. 4.2 COMPOSING THE PHASE DIAGRAMS For two of the catanionic systems, diphenhydramine/SDS and tetracaine/SDS, thorough investigations of the catanionic formation were performed and simplified ternary phase diagrams were composed (Paper I). Drug and SDS were weighed into vials and dissolved in 0.9% NaCl, varying the drug/SDS ratio and the total concentration of both. For the other systems investigated (Papers I – IV), the phase studies were not as thorough; the total concentration was kept constant at a maximum of two different concentrations, and only the drug:surfactant ratio was varied. The total concentration of drug and surfactant was never higher than 300 mM and generally the concentration studied was 40 mM. With the exception of the investigations on the influence of the ionic strength, all samples were dissolved in 0.9% NaCl. Similarly, the pH was not adjusted in any of the samples, other than those used when investigating the effects of the pH, where the pH was adjusted using NaOH and HCl. Upon mixing, the samples were allowed to equilibrate for at least a week, to minimize any effects introduced by differences in the preparation procedure, and to allow phase separation to occur where applicable. The mixtures were then studied by at least one of three different methods: visual inspection, rheological measurements and cryo-TEM. The most simple of the methods, that of visual inspection, proved to be very reliable. Obviously, precipitates and phase separation are easily spotted. Likewise, vesicles are easily recognizable by their opacity and for having a grey to slightly bluish color. Samples containing micelles are transparent, but large micelles can, however, still be recognized by the increased viscosity they induce, an increase that is notable even by eye. For more detailed studies of a potential increase in the viscosity, rheological measurements were conducted, as described below. Cryo-TEM was used to picture the vesicular and micellar structures formed in the samples. A great advantage of the cryo-TEM method is that, in contrast to the visual inspections, it allows the phase behavior to be examined even when the samples are incorporated in gels. Details of the cryo-TEM method are described elsewhere [110]. All samples were kept for at least two months before another visual inspection took place to study the effects of storage on the mixtures.. 4.3 PREPARATION OF GELS All the gels used in this thesis were prepared at a final concentration of 1% polymer. When preparing the Agar gels, the reference drug or the catanionic mixture was first dissolved in 0.9% NaCl, whereupon 1% polymer 23.

(169) Experimental Section powder was dispersed in the solution. The dispersion was then heated in a water bath and maintained at 100 ºC for 20 minutes to enable the polymer powder to dissolve completely. The resulting agar solution was transferred to the gel containers used for the drug release experiments, where it cooled down to form a stiff gel. Two alternative methods were used for preparing the Carbopol gels, one where all substances were weighed directly and together and one where a double strength gel was mixed with a double strength solution. Using the former method, the reference drug or the catanionic mixture was dissolved in 0.9% NaCl. The polymer powder was then dispersed in the solution and stirred, using magnetic stirring bars, for approximately 1 hour before NaOH was added to obtain a pH of about 7. The gel was allowed to equilibrate overnight and, the following day, the pH was set to 7.4 ± 0.1. In the last step of the preparation, some additional 0.9% NaCl was added to the final volume. Using the alternative method for Carbopol gel preparation, a solution of reference drug or catanionic mixture was prepared at twice the final concentration in 0.9% NaCl, or in 0.45% NaCl on some occasions in Paper IV. Parallel to this solution a 2% Carbopol gel being prepared and set to a pH of 7.4 ± 0.1, or to 10 ± 0.1 as was used sometimes in Paper IV, in accordance with the method described above. The double-strength solution and gel were then mixed (1:1 w/w), leaving a 1% Carbopol gel with the proposed final concentration of reference drug or catanionic mixture. Finally, the pH was checked once again and, if necessary, adjusted to the intended pH.. 4.4 RHEOLOGICAL MEASUREMENTS Rheological measurements were performed both on gels and on catanionic mixtures in solution. All rheological measurements were conducted using a Bohlin VOR Rheometer [111] (Bohlin Reologi, Lund, Sweden). When investigating the gels, dynamic oscillating measurements were performed. The gel was transferred to a concentric cylinder (C14) for the measurement which, prior to the measurement, was centrifuged at 3000 rpm to remove entrapped air. A few droplets of silicon oil were added to the surface to avoid evaporation. The measurement was then performed at 20ºC. Initially, a strain sweep measurement was made on each sample to define the linear viscoelastic region, within which the dynamic oscillation measurement would be conducted. Thereafter, the elastic module (G’), viscous module (G’’) and the phase angle () were determined.. 24.

(170) Experimental Section For the solutions of the catanionic mixtures, viscosity measurements were carried out. Concentric cylinder systems were used for these measurements as well (C8 and C14). Each investigation started by determining the delay time and thereafter the actual measurement was performed.. 4.5 DRUG RELEASE STUDIES The drug release studies were performed using a modified USP paddle method. The set-up used was a Pharma Test PTW II USP bath (Pharma Test Apparatebau, Germany) incorporating six beakers, in which custom-made sample containers could be immersed. Two different types of custom-made container were used, one for gels (Papers I-VI) and one for solutions (Paper VI). The two types of container are shown in Figure 6. Screws. Screws. Top. Top. Stainless Steel Net. Dialysis Membrane O-ring. Plastic Net. Stainless Steel Ring. Sample Container (6 ml). Hole for filling the cell Sample Container (6 ml) O-ring Screw. Figure 6. The drug release sample containers used when studying the release from (left) gels and (right) catanionic mixtures in solution.. The container for the gel experiments was filled with the gel, which was then covered with a coarse mesh-size plastic net, with the purpose of hindering diffusion of the polymer, followed by a stainless steel net to prevent the gel from expanding or changing shape in the container, and the container was then screwed together with its top section. When using the container for the solution experiments, however, the container was joined up prior to filling it with sample, and the solution was then added through a hole in the bottom of the cell, which was then properly sealed. No nets were needed in this cell, but a dialysis membrane ((CE) Membrane Sheets, MWCO 3.500, Spectra/Por®Biotech, CA) ensured that the solution was unable to escape the container. A cut-off of 3500 Da was chosen for the molecular weight to enable the monomer drug and SDS to diffuse freely through the membrane, whereas the catanionic aggregates would stay trapped inside the container.. 25.

(171) Experimental Section The gel containers were immersed in 250-750 ml drug release medium, usually comprised of 0.9% NaCl. The volume used was dependent on the spectrophotometrical detection of the drug and was chosen so that the absorbance remained at detectable levels throughout the entire experiment; in addition, care was taken to make certain that all the release studies were performed under sink conditions. Where the influence of the pH was studied, the pH of the drug release medium was set to that of the gel investigated, but in general the pH of the receiving medium was not adjusted. With the help of a peristaltic pump and ismaprene tubing (Ismatec SA, Zürich, Switzerland), the receiving medium was continuously pumped through a UV-vis spectrophotometer (Shimadzu UV-1601, Shimadzu, Kyoto Japan). The absorbance was measured automatically at regular intervals and stored using IDIS tablet dissolution data management software (Icalis Data Systems Ltd, United Kingdom). The experiments were generally performed in triplicates and the absorbance data was used to approximate the Fickian diffusion coefficient, D, for each experiment, using Equation 1.. Q. 2C 0 (. Dt. S. )1 / 2. (1). where Q is the amount of drug released per unit area, C0 is the initial concentration of drug in the gel and t is the time lapse since the experiment commenced. The equation is valid for approximately the first 60% of the fractional release [112, 113], which was taken into consideration whenever calculating the diffusion coefficient, and usually meant that the calculations were based on the first 40 minutes of the release. In Paper VI, not only was the drug release measured, but also the release of SDS. However, as SDS cannot be detected using a spectrophotometer, an alternative method had to be applied, Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES). Samples of the receiving medium were collected continuously throughout the experiments and stored in a freezer at -20 ºC before being analyzed. The instrument used was a Spectroflame P ICP-AES (Spectro Analytical Instruments, Kleve, Germany), equipped with a standard Meinhard nebulizer. Sulphur was monitored at the 180.734 nm atomic emission line. Details of the instrumentation and its use are described elsewhere [114].. 26.

(172) Experimental Section. 4.6 DETERMINATION OF THE CMC Solutions of each surfactant by itself, as well as solutions of mixed surfactants were prepared using 0.9% NaCl. The critical micelle concentration (CMC) of each solution was determined at room temperature, using the drop-weight technique on a custom-built instrument reproduced from Tornberg [115]. A vertically fixed syringe was filled with solutions containing the surface active substance under examination in various concentrations. A step motor drove a piston on the syringe, to induce the formation of a drop on the tip of the syringe. When the drop became detached from the tip it passed a photocell causing a signal to be registered on a recorder. The step motor was connected to an oscillator and the number of pulses that arose between the detachment of each drop could therefore be recorded. This number is proportional to the volume of the drop, which, in turn, is proportional to the surface tension of the solution, so the CMC can be determined by plotting the number of pulses as a function of the concentration of the solution. For each concentration of every solution, 9 to 11 measurements were made and the median values were used. All measurements were performed at the water/air interface. Details of the method are described elsewhere [116].. 4.7 THE INTERACTION PARAMETER According to the regular solution theory model for mixed micelles of two surfactants, described by Holland and Rubingh [117], the monomer concentration, Ci, is given by Equation 2. Ci. xi CMCi e E (1 xi ). 2. (2) where CMCi is the CMC of the surfactant i, xi is the mole fraction in the mixed micelles for the surfactant i and the interaction parameter, , can be determined using the mixed CMC, CMCmix, obtained for the catanionic mixtures, according to equation 3.. ln>D i CMC mix /( xi CMCi )@ (3) (1 xi ) 2 where Di is the mole fraction of surfactant i in the bulk solution and xi, must. E. be determined iteratively, using equation 4.. ªD CMC mix º 2 x1 ln « 1 » ¬ x1C1 ¼. ª D C º (1 x1 ) 2 ln « 2 mix » ¬ (1 x1 )C 2 ¼. (4). 27.

(173) Experimental Section The -parameters determined could then be used for anticipating the release of the catanionic components from gels. Description of the modeling and the numerical analysis used is beyond the scope of this thesis. Further information can be found in Paper VI.. 4.8 STATISTICAL ANALYSIS All drug release studies were performed in triplicate and the mean, the standard deviation (s.d.) and the 95% confidence interval (CI) were calculated for the diffusion coefficients for each triplicate. ANOVA was used, with Bonferroni’s multiple comparison test as post hoc test. The software used was Prism 4 for Windows by GraphPad Software Inc. (San Diego, CA).. 28.

(174) Results and Discussion. 5 RESULTS AND DISCUSSION. The work conducted for this thesis has evolved naturally over time. In these sections, the results of the publications included will be summarized briefly, although not necessarily paper by paper, but rather, according to the order in which the ideas popped up during the project. Before work commenced, in a study published by Paulsson and Edsman, it had been discovered that mixtures of amphiphilic drugs and oppositely charged surfactants formed vesicles [118]. In addition, these vesicles were able to slow the release of the drug in question, when released from gels. And so, the scene was set for the work presented in this thesis.. 5.1 DO DRUGS FORM CATANIONIC AGGREGATES WITH OPPOSITELY CHARGED SURFACTANTS? In Paper I, mixtures of SDS and three different cationic drugs were studied. The drugs used were diphenhydramine and tetracaine, for which extensive, though simplified, ternary phase diagrams were composed (see Figure 7), and amitriptyline, for which a less extensive phase study was performed.. Figure 7. Phase diagrams redrawn from Paper II. The white area symbolizes precipitates, the grey ones symbolize micellar/aqueous solution, the black areas symbolize vesicles and the striped ones represent multi-phase regions. The pH was left unadjusted when constructing both diagrams. The left phase diagram shows the three-component system containing diphenhydramine, SDS and physiological sodium chloride solution and the right one shows the system containing tetracaine, SDS and physiological sodium chloride solution.. 29.

(175) Results and Discussion Although the diphenhydramine/SDS and the tetracaine/SDS system resemble each other to a certain extent, there is one large difference: the diphenhydramine/SDS system has a vesicle area only on the SDS rich side, whereas the tetracaine/SDS system has one vesicle area on each side of the equimolar line. In Paper I this was interpreted as being indicative of that the tetracaine vesicles are able to exist with both a positive and a negative net charge, but as we shall see later in this thesis, this might not be the case. In both systems, however, vesicles and large micelles were found, as was also the case in the amitriptyline/SDS system, where a smaller study was conducted, varying only the ratio of drug and SDS while keeping the total concentration constant. Vesicles and micelles characteristic for most of the catanionic drug-surfactant systems examined are shown in Figure 8.. Figure 8. Cryo-TEM micrographs of (left) vesicles in a 14/26 mM tetracaine/SDS solution and (right) micelles in an 8/32 mM diphenhydramine/SDS solution. The bars show 200 nm.. Interestingly, the micrographs on the micelles do not only show elongated micelles, but also branched ones, as indicated by the arrows in Figure 8. However, cryo-TEM is an excellent method for qualitatively studying the appearance of the different structures present, but not quite so good when it comes to quantifying the structures. A rheological study was, therefore, performed, as shown in Figure 9.. 30.

(176) Results and Discussion. Figure 9. The viscosity of some of the compositions examined in the diphenhydramine/SDS system, reproduced from Paper I. The concentration SDS is plotted along the x-axis, whereas the corresponding diphenhydramine concentration is specified for each sample examined. Filled symbols represent vesicle phase and open symbols represent micellar/aqueous solution.. It has previously been shown that rheology may be used as a tool for indicating differences in the size and/or number of catanionic aggregates formed [119-121]. Studying Figure 9 it is possible, for example, to see how the micelles formed on the SDS rich side become smaller with an increasing fraction SDS as the viscosity decreases until it finally reaches a value close to that of water (about 1 mPa·s). In Paper II, the number of drugs tested for catanionic formation was increased by six, to a total number of nine. The selection criterion for this study was that the drugs should have either a positive or a negative charge and that their structure should “look” as though it would be surface active when going through the Swedish pharmaceutical specialties formula compendium [122]. The results from this study showed that both positively and negatively charged drug compounds were able to form catanionic mixtures when mixed with different oppositely charged surfactants (see Table 1).. 31.

(177) Results and Discussion Table 1. The occurrence of catanionic interactions tabulated according to the system examined in Papers I & II. Paper Cationic component. I I I II II II II II II II II II II II II II. Diphenhydramine Tetracaine Amitriptyline LPC TTAB BAC LPC BAC Diphenhydramine Diphenhydramine Lidocaine Orphenadrine Alprenolol Propranolol TTAB BAC. Catanionic Catanionic Catanionic Anionic interactions vesicles high-viscosity component micelles. X X X. X X X X X X X X X X X. X X X. X X X. X X X. X X X X. SDS SDS SDS Naproxen Naproxen Ibuprofen Ibuprofen Naproxen Naproxen Ibuprofen SDS SDS SDS SDS Ibuprofen SDS. For the prolonged release application intended (see Section 5.2), the large aggregates were of specific interest. Thus, Table 1 particularly concentrates on where vesicles and high-viscosity micelles were found, high-viscosity micelles being large ones. However, it should be noted that although large aggregates were not found in all systems examined, all but two systems exhibited some sort of interaction. Of the nine drugs examined in Papers I & II, six were able to form catanionic vesicles and seven were able to form large micelles when mixed with oppositely charged surfactants. Only one of the drugs, naproxen, formed neither vesicles nor large micelles in any of the compositions used. The question posed was whether drugs do form catanionic aggregates with oppositely charged surfactants, and evidently the answer to that question is: They do indeed. All nine drugs examined participated in interactions with oppositely charged surfactants and eight of them formed large catanionic aggregates. Considering the simplified selection criteria, where drugs were picked based merely upon the appearance of their structures, it is probably safe to assume that many positively or negatively charged drugs with surface active qualities are able to participate in the formation of catanionic mixtures with oppositely charged surfactants. As shown below, this quality may be used to benefit a prolonged release from gels, but the poten32.

(178) Results and Discussion tial negative effects must also be appreciated when working with other types of formulations. In most of the experiments, SDS was used, this is a surface active agent not uncommon in other formulations. Similarly, the commonly used preservative benzalkonium chloride (BAC) also proved to be active in catanionic interactions with both naproxen and ibuprofen. Hence, there is a risk of unwanted interactions, e.g. precipitation, or of unforeseen effects on the drug release, that could potentially affect the formulation in an undesirable manner.. 5.2 DO CATANIONIC AGGREGATES AFFECT THE DRUG RELEASE FROM GELS? In Paper I several compositions of two different catanionic drugsurfactant systems were tested to examine the extent of the prolongation of the drug release and a further three systems were tested in Paper II. All five systems examined indicated that using catanionic systems is a very efficient method of obtaining prolonged release from gels. Some drug release results from Paper I are shown in Figure 10. From the diphenhydramine/SDS system, two vesicle and one micelle compositions were subjected to drug release trials. As described in Section 5.1, the tetracaine/SDS system was able to form vesicles on both sides of the equimolar line. Naturally, a sample from each vesicle region was released from gels. Unfortunately, however, from visual inspections and rheological results, it was obvious that the vesicle from the drug-rich side would not coexist well with the Carbopol gel. The Carbopol was therefore exchanged for Agar for some of the drug release studies. 1,2. 1,2. 1. 1 1. 0,8. Fraction Released. Fraction Released. 1. 2. 0,6. 3 0,4 0,2. 2 0,8. 3 0,6. 0,4. 4. 0,2. 5. 0. 0. 0. 50. 100. 150. 200. 250. Time (min). 300. 350. 400. 450. 0. 50. 100. 150. 200. 250. 300. 350. 400. 450. Time (min). Figure 10. An example of drug release profiles. To the left is the diphenhydramine/SDS system, where (1) marks the 14 mM diphenhydramine reference in 1% C940, (2) is the 8/32 mM micellar composition in 1% C940 and (3) is the 14/26 mM vesicular composition in 1% C940. To the right is the tetracaine/SDS system, where (1) marks the 14 mM tetracaine reference in 1% Agar, (2) is 14 mM tetracaine reference in 1% C940, (3) is the 26/14 mM vesicular composition in 1% Agar, (4) is the 14/26 mM vesicular composition in 1% C940 and (5) is the 14/26 mM vesicular composition in 1% Agar.. 33.

(179) Results and Discussion The release of diphenhydramine from the Carbopol gel was significantly slower when using catanionic micelles or vesicles, than when using diphenhydramine alone; the diffusion coefficient was at least 10 times slower for both of the catanionic compositions tried. Interestingly, however, there was no difference in the drug release rate when comparing micelles with vesicles, supporting the evidence obtained from rheological investigations that the micelles constitute quite large structures. The tetracaine release from the Carbopol gel, using the vesicle from the SDS-rich side, was delayed to a greater extent than the diphenhydramine release using the SDS-rich diphenhydramine vesicle; the diffusion coefficient was about 60 times smaller than that of the reference gel, containing tetracaine alone. The release from an Agar gel using the vesicle from the tetracaine-rich side was prolonged as well, compared to the reference gel, although not as much as when using the vesicle from the SDS-rich side. The diffusion coefficient was only about five times smaller for the vesicle composition than for tetracaine alone. The reason why the release from the two different tetracaine vesicle compositions differed was probably not the gels, despite the fact that different gels were used. A comparative study was performed, measuring the drug release using the SDS-rich tetracaine vesicle from both Carbopol and Agar gels, and, although the diffusion coefficient differed for the two gels, the difference was much smaller than it was between the two different vesicle compositions. Rheological measurements were performed to investigate to identify any catanionic effects on the Carbopol gels. A large change in the rheology of the gel could indicate interactions with any of the catanionic components, which, themselves, could affect the release from the gel. In addition, a change in rheology might affect the mucoadhesive properties of the gel. Fortunately, however, only small changes in rheology were recorded. CryoTEM was also used to control that the aggregates were still present once they had been incorporated into the gel. The micrographs, some of which are shown in Figure 11, confirmed that both the vesicles and the micelles seemed unaffected by the presence of the polymer.. 34.

(180) Results and Discussion. Figure 11. Cryo-TEM micrographs of (left) vesicles in a 14/26 mM tetracaine/SDS mixture and (right) micelles in an 8/32 mM diphenhydramine/SDS mixture, both in 1% C940. The bar indicates 200 nm.. In Paper II, three new catanionic systems were used in drug release studies: lidocaine/SDS, orphenadrine/SDS and ibuprofen/TTAB, the lidocaine and orphenadrine being positively charged and the ibuprofen being negatively charged. Two of the mixtures, orphenadrine/SDS and ibuprofen/TTAB, were released from Agar gels owing to incompatibilities when mixed with the Carbopol gel, whereas the lidocaine/SDS system mixed flawlessly with the Carbopol. The results supported those from Paper I, the catanionic mixtures enabled diffusion coefficients to be obtained of 10-100 times smaller than those from the reference gels containing the drug substance alone. A demonstration of the diffusion coefficients for each mixture examined in Papers I & II, compared with its corresponding reference, is shown in Figure 12. Within the lidocaine/SDS system, some investigations concerning the dependence of the total concentration of the drug and surfactant together were performed. Keeping the lidocaine/SDS ratio constant for two micellar compositions while varying the total concentration between 40 and 160 mM illustrated that varying the total concentration did not affect the drug release rate. For both fractions in which the total concentration was varied, the diffusion coefficient obtained was more or less the same at all total concentrations. In addition, the drug release rate was about the same when varying the lidocaine/SDS ratio as well. All in all, seven different micellar lidocaine/SDS compositions were studied and any difference in the diffusion coefficients was non-significant. As for the ibuprofen/TTAB system, the drug release results were similar to those for the lidocaine/SDS system, with no significant differences in the 35.

(181) Results and Discussion drug release rate being noted when comparing three different micellar compositions. In a comparison with the reference gel containing ibuprofen alone, however, the diffusion coefficients were about 10 times smaller. 25 20 15 10 5. Di f/. Di f/. SD. S, 14 / SD 26 m S, M 8/ 32 m M Li d/ SD S, Li 8/ d/ 32 SD m S, M Li 1 d/ 2/ SD 28 S, m M Li 16 d/ /2 SD 4 m S, M Li 20 d/ /2 SD 0 S m Li ,2 M d/ 4/ SD 16 S, m 32 M Li d/ SD /12 8 S, m M 64 /9 6 m M TT AB /I b TT u, AB 8/ 32 / Ib m u, TT M 12 AB / 2 / Ib 8 m u, M 16 /2 4 m M. 0. 300. 250 200. 150 100. 50 0 Tet/SDS, 14/26 mM. Tet/SDS, 14/26 mM. Tet/SDS, 26/14 mM. Orph/SDS, 16/24 mM. Orph/SDS, 32/128 mM. Orph/SDS, 48/112 mM. Figure 12. An illustration of the extent to which the drug release was slowed for the different catanionic mixtures. Each column shows the reference diffusion coefficient (i.e., when the drug was not mixed with an oppositely charged surfactant) divided by the diffusion coefficient from the catanionic mixture indicated on the xaxis and therefore provides a direct comparison of the relative increase in the time over which the drug is released. The bars show the confidence interval for each comparison. In the orphenadrine/SDS system, both vesicles and micelles were subjected to drug release. Comparing the diffusion coefficients for the micelle with that for the drug reference, there was a ten fold difference. However, a comparison of the diffusion coefficient for the two vesicles examined with the same reference revealed an astounding 100-fold difference in the diffusion coefficients for one of the vesicles and a 50-fold difference for the other.. 36.

(182) Results and Discussion As stated above, the rheology of the Carbopol gels used in the release studies in both Papers I & II was investigated, to ensure that the gel was unaffected by the incorporation of the catanionic components. It was previously discovered that some of the systems would not mix well with the Carbopol, whereupon Agar was used instead, but for the systems released from Carbopol, the rheological results indicated no major interactions between the catanionic mixture and the polymer, and definitely not on a level that would explain any part of the prolonged release. To conclude this section, it is obvious that using catanionic mixtures is a very efficient method of obtaining prolonged release from gels. The diffusion coefficient can be decreased by a factor of 10 to 100 compared with the diffusion coefficient for a gel containing drug substance alone. The mechanism suggested for the prolonged release was that the catanionic micelles and vesicles were too large to move inside the gel, and that the diffusion rate therefore was dependant on the fraction drug not involved in formation of these large aggregates. One interesting point should be made though, namely, in Paper I it was discovered that the tetracaine/SDS vesicles from the tetracaine-rich side of the phase diagram would not mix well with the Carbopol gel, and this poor compatibility was probably slightly misinterpreted in Paper I as it turned out from the follow-up study, summarized in the next section.. 5.3 WILL THE USE OF CATANIONIC MIXTURES WORK IN A PHYSIOLOGICAL ENVIRONMENT? Already in Paper I some pilot studies on the influence of the pH and ionic strength were performed on a diphenhydramine/SDS vesicle. The pH in the catanionic mixture was adjusted between 0.5 and 11 and visual confirmation revealed that it was not until the pH reached 11 that any change was observed in the samples, i.e. two units above the pKa of diphenhydramine. At pH 11, the characteristic bluish-grey opacity of a vesicle phase diminished and the sample turned completely transparent, signifying a vesicle to micelle shift. The diphenhydramine/SDS vesicle composition was also subjected to drug release studies having raised the pH of the gel, as well as the release medium, to 11.7, and it was concluded that although the drug release was faster than in an experiment conducted at pH=7.4, the drug release was still significantly prolonged compared to the release for a reference gel. When adjusting the ionic strength in the same diphenhydramine/SDS vesicle composition the mixture seemed quite resilient to the changes in NaCl concentration. The ionic strength was increased, raising the NaCl con37.

(183) Results and Discussion centration from 0.9% to 5.4% NaCl, and it was not until the concentration reached 3.6% (i.e., four times the physiological osmolality) that the sample showed any change that was visible in a visual inspection. Consequently, the results from Paper I indicated that the catanionic mixtures can tolerate quite large environmental variances. However, a more thorough investigation on the influence of changes in ionic strength and pH was necessary, and thus the studies reported in Paper III were performed. In Paper III, the two systems investigated most thoroughly in Paper I, diphenhydramine/SDS and tetracaine/SDS, were studied again. This time, however, the entire examination was devoted to determining how they would react to changes in the pH and ionic strength. The emphasis was placed on the diphenhydramine/SDS system. An initial phase study was performed, keeping the total concentration of drug and surfactant constant at 40 mM, while varying the drug/surfactant ratio. The pKa values of diphenhydramine and tetracaine are 9.0 and 8.5, respectively, and one issue of interest in Paper III was the effect of having a pH above the pKa of the cation. Thus, the changes in the pH examined spanned the range from neutral to 10 for the tetracaine system and to 11 for the diphenhydramine system. Increasing the pH to 10 or 11 may not be pharmaceutically relevant for direct usage, but as it is of interest to learn about the effects of a pH value close to the pKa of the cation, these high pH values were included in the study. In this respect, the systems investigated can be regarded as model systems. The variations in the ionic strength were, however, restricted to what were considered to be pharmaceutically relevant concentrations of NaCl, 0.45% to 1.8%. The results from Paper III support the conclusion from Paper I to a great extent, confirming that the catanionic aggregates were quite resilient to changes in pH, and even more impervious to changes in the ionic strength. The phase studies performed in the diphenhydramine/SDS system revealed that doubling the NaCl concentration from 0.9 to 1.8 NaCl brought about no changes. Brahser et al have shown how an increased concentration of NaBr eventually made the vesicles formed in a CTAB/SOS system turn into micelles [104]. Similar effects had already been spotted in the diphenhydramine/SDS system as reported in Paper I, where a vesicle-to-micelle transformation occurred at 5.4% NaCl, although such an ionic strength is probably pharmaceutically irrelevant. Furthermore, only in one out of nine samples was a phase shift observed when the concentration of NaCl was halved to 0.45%. Variations in the pH turned out to influence the systems more than variations in the ionic strength. In the diphenhydramine system, both the micellar and the vesicle phase regions seemed to prevail, and possibly even grow, as the pH was increased. This was surprising as the effects were seen even 38.

(184) Results and Discussion when the pH was raised to above the pKa of diphenhydramine. However, there was put forward a conceivable explanation for this. The compositions investigated were on the SDS-rich side, meaning that the aggregates probably had a negative net charge. (A supposition that was supported in a subsequent study, see Paper VI.). A net negative charge would attract hydrogen ions, resulting in a local pH that is lower than that observed on a macroscopic level. At pH 10 and 11, vesicle formation occurred, even at an equimolar drug/surfactant ratio, and the micellar phase area extended further towards an increased fraction SDS. The explanation put forward was that the fraction of charged drug compound decreased with increasing pH. The fraction available for catanionic formation would, therefore, decrease which would show as a phase shift towards an increased SDS fraction. Rheology was used to investigate the influence of the pH on the size and/or number of micelles, which ought to be proportional to the viscosity of the sample. The results are shown in Figure 13. The viscosity of the different samples confirmed what had been revealed in Paper I, that an increased total concentration increases the micellar size. In addition, it also indicates that the viscosity is inversely proportional to the pH, meaning that increasing the pH in a micellar solution will cause the micelles to become smaller. This is in accordance with the phase shift related to the lesser fraction of charged drug compound, described above, i.e., increasing the pH will leave a smaller amount of charged diphenhydramine free to participate in catanionic formation, and there will be a phase shift corresponding to an increased fraction of SDS in the phase diagram.. Viscosity (mPa s). 8 1.. 7 6 2.. 5 4 3 2. 3.. 1 0 7. 8. 9. 10. 11. 12. pH Figure 13. An illustration of how the viscosity of catanionic micellar mixtures is affected by changes in pH. The three different compositions are 1. 12 mM diphenhydramine and 28 mM SDS, 2. 8 mM diphenhydramine and 32 mM SDS and 3. 4 mM diphenhydramine and 36 mM SDS. 39.

(185) Results and Discussion Although the pH had an influence on the phase behaviour of the diphenhydramine/SDS system, both vesicles and micelles were observed at all the values of pH investigated, and increasing the pH to above the pKa of the drug even seemed to extend the areas for some interesting phases. Unfortunately, the results for the tetracaine/SDS system were not quite as promising. Already at a pH of 8, the tetracaine rich vesicle phase had disappeared, and increasing the pH to 10 had obvious effects on the SDS-rich vesicle. However, as Figure 14 demonstrates, the vesicle phase could still be found at that pH, although it was less extensive.. Figure 14. Cryo-TEM micrographs at pH=10 of (left) vesicles in a 14/26 mM tetracaine/SDS solution and (right) micelles in an 8/32 mM diphenhydramine/SDS solution. The bar indicates 200 nm.. The results from this phase study also allow the correction of a previous assumption, made in Paper I, where it was thought that the reason why the tetracaine-rich vesicle would not mix with the Carbopol gel was an interaction between the vesicle and the polymer. A more probable reason now seems to be that the effect was associated with the pH, considering that the Carbopol gel was always adjusted to pH=7.4 prior to conducting the release studies, and considering how the phase study in Paper III shows that the tetracaine vesicle disappeared when the pH was increased. From the drug release studies it can be concluded that drug release using catanionic vesicles is only slightly affected by changes in either the pH or the ionic strength — as long as there is no vesicle-to-micelle transition. Lowering the NaCl concentration to 0.45% did not affect the drug release from a diphenhydramine/SDS vesicle at all, and increasing the pH to 10 for the same composition only caused a small, and non-significant increase in the diffusion coefficient (see Figure 15).. 40.

(186) Results and Discussion 80 70 60 50 40 30 20. μ pH pH. 10. μ. pH. 0 Dif/SDS, Dif/SDS, Dif/SDS, Dif/SDS, Dif/SDS, Dif/SDS, 14/26 mM 14/26 mM 14/26 mM 8/32 mM 8/32 mM 8/32 mM. Tet/SDS, Tet/SDS, 14/26 14/26. Figure 15. An illustration of the extent to which the drug release was slowed for the different catanionic mixtures. Each column shows the reference diffusion coefficient (i.e., when the drug was not mixed with an oppositely charged surfactant) divided by the diffusion coefficient from the catanionic mixture indicated on the x-axis and therefore provides a direct comparison of the relative increase in the time over which the drug is released. When nothing else is indicated the experiment was performed at pH=7.4 and in 0.9% NaCl, but the μ-note means that the NaCl concentration was 0.45% and the pH-note means that the experiment was performed at pH=10. The bars show the confidence interval for each comparison.. For the diphenhydramine/SDS micellar sample, however, the diffusion coefficient was influenced by changes in both the pH and the ionic strength. As shown in Figure 13, the micelles grow smaller when the pH is increased. The smaller the micelles, the lower the resistance to movement through the gel, which could explain the increased diffusion coefficient obtained at at higher levels of pH. Such pH-related effects were not observed in the vesicle studies, either because the vesicle size was not affected by changes in the pH or because the vesicles were sufficiently large to be immobile even if there were an effect on the vesicle size. Lowering the ionic strength had a similar effect on the viscosity of the samples as increasing the pH. The viscosity dropped from 6 to about 2 mPas when the NaCl concentration was lowered from 0.9% to 0.45%, indicating that the micelles were smaller, which in turn would explain the increased release rate at the lower ionic strength. Rheological investigations on the gels used in the drug release studies revealed that only small changes occurred in the rheology of any of those used and confirmed that all the gels fulfilled the rheological definition of a gel regardless of the variations in pH and ionic strength. The ionic strength is probably the most likely variable to influence the rheology, rather than the pH, because an increased ionic strength may screen the repulsive forces contributing to the swelling of the gel [123-125]. The strength of the gel should,. 41.

No results found